Anthranilate synthase (AS) catalyzes the first reaction branching from the aromatic amino acid pathway to the biosynthesis of tryptophan in plants, fungi, and bacteria. The reaction catalyzed by anthranilate synthase is the conversion of chorismate to anthranilate in a glutamine-dependent reaction. In microorganisms, anthranilate synthase is composed of two non-identical subunits: the alpha subunit binds chorismate and eliminates the enolpyruvate side chain, and the beta subunit transfers an amino group from glutamine to the position vacated by the enolpyruvate moiety.
The next reaction in the synthesis of tryptophan is the transfer of the phosphoribosyl moiety of phosphoribosyl pyrophosphate to anthranilate. The indole ring is formed in two steps involving an isomerization converting the ribose group to a ribulose followed by a cyclization reaction to yield indole glycerol phosphate. The final reaction in the pathway is catalyzed by a single enzyme that may contain either one or two subunits. The reaction accomplishes the cleavage of indole glyceraldehyde-3-phosphate and condensation of the indole group with serine (Umbarger, Ann. Rev. Biochem, 47, 555 (1978)).
Metabolite flow in the tryptophan pathway in higher plants and microorganisms is apparently regulated through feedback inhibition of anthranilate synthase by tryptophan. Thus, because anthranilate synthase is feedback inhibited by tryptophan, the overproduction of wild-type anthranilate synthase cannot result in tryptophan overproduction.
While anthranilate synthase has been partially purified from crude extracts of cell cultures of higher plants (Hankins et al., Plant Physiol., 57, 101 (1976); Widholm, Biochim. Biophys. Acta, 320, 217 (1973)), it was found to be very unstable. In order to further characterize the anthranilate synthase of plants, Niyogi and Fink (Plant Cell, 4, 721 (1992)) and Niyogi et al. (Plant Cell, 5, 1011 (1993)) employed a molecular approach. They found that Arabidopsis anthranilate synthase alpha subunits are encoded by two closely related, nonallelic genes which are differentially regulated. One of these alpha subunit genes, ASA1, is induced by wounding and bacterial pathogen infiltration, implicating its involvement in a defense response, whereas the other alpha subunit gene, ASA2, is expressed at constitutive basal levels. Both predicted proteins share regions of homology with bacterial and fungal anthranilate synthase proteins, and contain conserved amino acid residues at positions that have been shown to be involved in tryptophan feedback inhibition in bacteria (Caligiuri et al., J. Biol. Chem., 266, 8328 (1991)).
Amino acid analogs of tryptophan or of intermediates in the tryptophan biosynthetic pathway (e.g., 5-methyltryptophan, 4-methyltryptophan, 5-fluorotryptophan, 5-hydroxytryptophan, 7-azatryptophan, 3.beta.-indoleacrylic acid, 3-methylanthranilic acid) have been shown to inhibit the growth of both prokaryotic and eukaryotic cultures. Plant cell cultures can be selected for resistance to these amino acid analogs. For example, cultured tobacco, carrot, potato, corn and Datura innoxia cell lines have been selected which are resistant to growth inhibition by 5-methyltryptophan (5-MT), an amino acid analog of tryptophan, due to expression of an altered anthranilate synthase as described below.
Widholm (Biochem. Biophys. Acta, 261, 44 (1972)) demonstrated that the tryptophan analogs 5-MT, 4-methyltryptophan, 5-fluorotryptophan and 6-fluorotryptophan cause growth inhibition of tobacco (Nicotiana tabacum) and carrot (Daucus carota) cell cultures. This inhibition of growth could be reversed by the addition of anthranilic acid, indole, or L-tryptophan. Anthranilate synthase was determined to be very sensitive to these analogs. The tryptophan analogs inhibited cell growth by limiting tryptophan synthesis through the inhibition of anthranilate synthase.
While growth of many cultured tobacco cell lines was inhibited by 5-MT, some tobacco cell lines were resistant to growth inhibitory concentrations of 5-MT (Widholm, Biochim. Biophys. Acta, 261, 52 (1972)). The resistant phenotype was stable for at least 60 cell mass doublings even without selection pressure (i.e., without 5-MT). In addition, 5-MT resistant cells were resistant to growth inhibition by other tryptophan analogs. Free tryptophan levels were increased in 5-MT resistant cells about 10-fold over control tissue. Anthranilate synthase in these 5-MT resistant cells was found to be less sensitive to inhibition by tryptophan or 5-MT.
Carrot cell lines that were resistant to growth inhibition by 5-MT were also selected by Widholm (Biochim. Biophys. Acta, 279, 48 (1972)). This phenotype was generally stable in the absence of the tryptophan analog for at least 100 cell doublings. 5-MT resistant cells were also resistant to other tryptophan analogs. Free tryptophan concentrations in 5-MT resistant cells were substantially increased to 2170 .mu.M as compared to 81 .mu.M (27-fold) for control tissue. Anthranilate synthase was shown to be altered in the 5-MT resistant cells. The enzyme was about 5-fold less sensitive to inhibition by tryptophan or 5-MT than an unaltered anthranilate synthase.
Singh et al. (Biochem. Genet., 13, 357 (1975)) described a mutant in corn, Zea mays L., blue fluorescent-1, that possessed increased anthranilate synthase activity which was less sensitive to feedback inhibition. The mutant also accumulated anthranilic acid. In contrast to previous work in tobacco and carrot, however, the altered anthranilate synthase activity did not lead to significant overproduction of tryptophan in mature corn plants or seed.
Hibberd et al. (U.S. Pat. No. 4,581,847, issued Apr. 15, 1986) described 5-MT resistant maize cell lines that contained an anthranilate synthase that was less sensitive to feedback inhibition than wild-type anthranilate synthase. One 5-MT resistant cell line accumulated free tryptophan at levels almost twenty-fold greater than that of non-transformed cell lines.
Carlson et al. (Physiol. Plant, 44, 251 (1978)) obtained potato cell (Solanum tuberosum) cultures resistant to 5-MT. Anthranilate synthase in these cultures was shown to be less sensitive to inhibition by tryptophan or by 5-MT, although both 5-MT resistant and sensitive forms of the enzyme were present in the cells of the culture. In the selected cell lines, the level of resistant anthranilate synthase was greatly increased relative to the level of the sensitive form. The range of free tryptophan concentrations in selected cultures was from 970 to 1400 .mu.M compared to control cultures in which the tryptophan concentrations were about 29 .mu.M.
Widholm (Plant Cell Cultures: Results and Perspectives, F. Sala, B. Parisi, R. Cella, O. Ciferri (eds.), Elsevier/North Holland Biomedical Press, Amsterdam, pp. 157-159 (1980)) described plants regenerated from 5-MT resistant N. tabacum suspension cultures. While the cultures expressed an anthranilate synthase enzyme that was less sensitive to feedback inhibition by tryptophan and also exhibited an increased level of free tryptophan (approximately 25-fold), the leaves of the regenerated plants did not express the altered form of the enzyme and did not form roots in medium containing 5-MT. The resistance trait was, however, expressed in callus derived from the regenerated plant. Thus it appears to be difficult to obtain expression of the 5-MT resistance phenotype in tobacco plants derived from 5-MT resistant cells selected in culture.
Finally, Ranch et al. (Plant Physiol., 71, 136 (1983)) selected for 5-MT resistance in cell cultures of Datura innoxia, a dicot weed, and reported that the resistant cell cultures contained increased tryptophan levels (8 to 30 times higher than the wild type level) and an anthranilate synthase with less sensitivity to tryptophan feedback inhibition. Regenerated plants were also resistant to 5-MT, contained an altered anthranilate synthase, and had a greater concentration of free tryptophan (4 to 44 times) in the leaves than in the leaves of the control plants. In contrast to the studies with N. tabacum, where the altered enzyme was not expressed in plants regenerated from resistant cell lines, these results indicated that the amino acid overproduction phenotype could be selected at the cellular level and expressed in whole plants regenerated from the selected cells in Datura innoxia.
Although it is possible to select for 5-MT resistance in certain cell cultures and plants, this characteristic does not necessarily correlate with the overproduction of free tryptophan in whole plants. Additionally, plants regenerated from 5-MT resistant lines frequently do not express an altered form of the enzyme. Nor is it predictable that this characteristic will be stable over a period of time and will be passed along as a heritable trait.
Thus, there is a need to increase the tryptophan content of plants and/or provide plants that are resistant to growth inhibitory amounts of tryptophan or an analog thereof.